Introduction

The study of the space between the state and the market (referred as ‘the space’ or ‘in-between’ space throughout the article, following Knutsen and Brock 2014)Footnote 1 remains under-theorized (Taylor 2010; Corry 2010). Most prominently, researchers have not agreed on the definition of the organizations in the space, or on the boundary of the study (e.g. the third sector, civil society, the non-profit sector).

Recently, scholars took note of a possible ‘paradigm’ change of the field of study prompted by debates on these definitions (Van Til 2009; Kramer 2000; Wagner 2012; Evers 2013). Most writers distinguished an America-led, non-distribution, constraint-based positivist non-profit ‘sector’ paradigm from a ‘new’ paradigm, emphasizing the blurry sectoral boundary, voluntariness, and normative values. Van Til (2009) describes that the field has been dominated by ‘a paradigm’ that focuses on organizations defined by their tax-exemption status via the non-distribution constraint. He argues that, although this paradigm is able to operationalize a ‘non-profit sector’, it potentially downplays other organizations in the space. Kramer (2000) notes a questioning of the ownership-based non-profit ‘sector’ concept and the emerging emphasis on the blurring sectoral boundaries. Wagner (2012) distinguishes an ‘economic paradigm’ that is primarily led by the United States non-profit empirical research from a ‘normative political paradigm’ that emphasizes voluntariness. He calls for the recognition of both paradigms and for more research about voluntariness. Additionally, Evers (2010, 2013) observed that civicness is missing from the US-led non-profit paradigm, and that more research is needed. In this article, we refer to the US-led paradigm with an economic focus as the ‘non-profit paradigm’. We refer to the emerging paradigm, which has a blurring sectoral boundary view as the ‘new paradigm’.

This article analyses these two strands of scholarship from an epistemological perspective. First, the article distinguishes between two kinds of empirical inquiries: inquiries under the non-profit paradigm, represented by delineating the ‘sector’, and inquiries under the new paradigm, represented by measuring hybrid organizations. The author argues that, although the new paradigm has never intended to provide an operationalization method for non-profit paradigm inquiries, this lack of methods contributes to the low visibility of the new paradigm. However, the lesson to learn is to develop operationalization methods to aid empirical work—not for the non-profit paradigm’s conventional inquiries, but for the new paradigm itself, such as measuring hybrid organizations. This article then introduces the organizational identity (OI) approach from organizational ecology as a potential method.

Metaphysical Theorization Versus Empirical Testing

The bifurcation of the scholarship has been observed over a decade ago. It seems to take a while for the ‘new paradigm’ to be known; there remains a relatively low level of visibility, particularly in the US, among empirical studies. Kramer (2000) observed two contrasting theoretical perspectives in the American study of the non-profit sector. One perspective ‘emphasizes the rapid institutionalization of a third sector as the primary partner with government’ (‘the non-profit paradigm’), and the other questions the ownership-based sectoral model and emphasizes the blurry sectoral boundary (‘the new paradigm’). Kramer points out that the second perspective is less well known. Similarly, Wagner (2012) also specifies two paths of scholarship. One path focuses on developing normative political theory, assuming a democratic society (‘the new paradigm’). The second path, with an economic empirical focus, concentrates on the decentralization of the administrative system and the partnership between non-profits and the government to provide services (‘the non-profit paradigm’). Wagner is concerned with the under-studied status of the normative political path and calls for more future research. Table 1 offers a brief contrast of the two paradigms and a summary for the discussions that follow.

Table 1 Contrasting two paradigms
Full size table

From a modern philosophical view, good science needs both metaphysical theorizing and empirical testing and, in fact, both are inseparable (Plotkin 1994; Alexander 1982). Alexander (1982) suggests viewing empiricism and metaphysical theorization as extremes on an epistemological continuum, and accepts that both are important to good science. Plotkin (1994) concludes that:

Science … proceeds by guessing at the nature of the world (theorizing) and then disciplining and revising those guesses by testing how they fit with the experienced world (observation and experimentation); in a sense, science combines rationalist and empiricist philosophies into an inseparable world (p. 19).

From this perspective, the non-profit paradigm largely takes on an empiricist orientation (Alexander 1982). The new paradigm takes on an orientation towards metaphysical theorization.

The non-profit paradigm is represented by the development of the structural–operational definition of non-profits (i.e. Salamon and Anheier 1997: private, non-profit distributing, self-governing, and voluntary), its follow-up works, and works applying this approach. The non-profit paradigm has been focused on what Taylor calls the ‘central task’ (Taylor 2010): ‘delineating the composition, scope, and structure of the third sector and to map the field, first nationally and then…cross-nationally and globally’ (1). Although this strand of work has produced plausible theories, measurable definitions, international data, and tested theories, it has been criticized for its lack of consideration of some crucial characteristics of the space, including the difficulty of drawing a sectoral boundary and a high level of diversity of the organizations in the space. These neglected characteristics challenge the concept of a discrete ‘third’ or ‘non-profit’ sector and challenge the validity of inquiries assuming a ‘sector’.

On the new paradigm side, although a more realistic account of the space has been prominently theorized, recognizing the blurring sectoral boundary and the hybrid organizations, this stream of work never intends to provide an operationalization tool to assist with the ‘central task’. Instead, the new paradigm rejects a ‘sector’ concept; therefore, it is not interested in delineating the sector. More attention is being paid on theorizing the blurring sectoral boundary view, developing new classifications of organizations, and conceptualizing hybrid organizations. However, when much scholarly attention was still directed at the central task, it would be difficult for the new paradigm to catch on quickly. Thus, this provides some explanation for the lack of popularity of the new paradigm in some areas of study. The next section reviews the key debates between the two paradigms. It further illustrates that the new paradigm has advanced the theorization of the in-between space, but it has not yet focused on operationalization tools for its own empirical research.

The New Paradigm: Advancing Theorization But a Lack of Operationalization Tools

The differences between the two paradigms can be derived from two fundamental assumptions that the non-profit paradigm seems to assume, and which the new paradigm rejects: (1) considering the space as a discrete ‘sector’, and (2) treating non-profits as the only organizations in the space. The non-profit paradigm has never formally espoused these assumptions, but these assumptions were mistakenly imposed on the non-profit paradigm, perhaps due to confusion and ambiguous use of the term “non-profit sector”.

Two Misleading Assumptions

Recently, Steven Rathgeb Smith (2011) offered a carefully drafted understanding of ‘non-profit organizations’. His description clarifies that (1) non-profits are only one category of organizations among many other organizations in the space, and that the space is very diverse; (2) the collectivity of non-profits should not be seen as the in-between space itself. Smith (2011) writes:

Civil society is the space of voluntary association and activity that exists in relative separation from the state and the market…. Non-profit organizations (or ‘non-profits’) form a specific category of associational life in civil society… Non-profits constitute an important element in the ecosystems of civil society, but the two should not be conflated (30).

However, in the practice of conducting research, reflecting the influence of the non-profit paradigm, there has been a tendency to perceive the in-between space as a coherent discrete ‘sector’ containing organizations that share common characteristics. Taylor (2010, p. 1) revealed that it is acceptable to use the ‘third sector’ as a ‘catch-all’ term to refer to all kinds of organizations in the space. The term ‘the third sector’ masks diversity; as asserted by Corry (2010): ‘The idea of third sector suggests that these entities, however diverse, together make up a coherent whole—a sector with its own distinct type of social norm and practical logic’ (11). Corry indicates that the term, ‘the third sector’, could have misled researchers to seek defining characteristics for this sector, which is indeed diverse and may not possess such characteristics. Corry writes that people want to ‘tidy up’ the sector: ‘if something is ruled neither primarily by market logic nor via a bureaucratic chain of command, it must be part of the ‘third’ sector’ (2010, p. 13).

A second assumption is to perceive that non-profits are the only organizations in the ‘third sector’. Perhaps, for convenience, equating the non-profit sector to the third sector is an easy step. This assumption is, at best, imprecise.

Rejections of the Two Assumptions: Early Voices on Boundary Blurring

The blurry sectoral boundary view is at the core of this debate (Buckingham 2011). It is the ‘real reason to challenge the tri-sectoral model of society and achieving international standardization of the field’ (Taylor 2010, p. 6). Accepting the blurry boundary view implies rejection of the aforementioned two assumptions, which are otherwise known as a ‘sector’ concept. The ‘sector’ concept is challenged by the fact that (1) organizations in the space are too diverse to be collectively considered a coherent ‘sector’; and (2) these organizations often internalize characteristics from other external sectors to a degree that they are de facto hybrid organizations.

In the early years of the field of study, there were voices that were already calling for recognition of the blurry boundary view. This idea emerged as early as the 1980s (special issue, Journal of Voluntary Action Research, 1987, Volume 16, Number 1–2). Van Til (1988), following Parsons (1966), observed that sectors conduct transactions with each other across boundaries. DiMaggio and Anheier (1990, p. 280) wrote: ‘the quest for generalizable differences among NPOs, proprietaries, and public agencies is problematic…lines between public, NP (not-for-profit), and FP (for-profit) enterprise are often unclear… we cannot assume that processes generating differences and similarities will persist far into the future’.

Alford (1992) argued that the ‘non-profit’ sector, which is distinguished from the ‘public’ and ‘private’ sectors, represents political language that conceals the actual relationship of some non-profits to the state and capital. He also noted that DiMaggio (1987), Langton (1987), Salamon (1987), Wuthnow (1991), Ware (1989), and Van Til (1988) all agreed that the tripartite distinction between public, private, and non-profit could be problematic because the distinction conceals the interrelationship among sectors. Billis (1993) reviewed the literature and concluded that the blurry boundary view is ubiquitous. Alford called for more attention to be paid to the non-profits’ relationship with the state and capital: ‘The imprecision of boundaries between different types of NPOs, in fact their overlaps, is part of the essential character of these organizations’ (1992, p. 40). Langton wrote, ‘segmented ties between non-profit institutions and government and business are both elastic and porous’ (1987, p. 143).

Studies began to emerge examining the boundary between the third sector and private or public sectors. Kramer (1984) focused on the interrelationships among private, governmental, and voluntary organizations and then identified five types of activities reflecting the interpenetration among sectors: reprivatization, empowerment, pragmatic partnership, government operation, and nationalization. Ferris (1993) recognized the threat to non-profit autonomy in partnering with government. Kramer and Grossman (1987) identified some negative effects on non-profits if the organization heavily relies on government, which included funding uncertainty, goal deflection, and reduced fiscal flexibility. Weisbrod (1998) predicted that non-profits would both compete and cooperate with private enterprises and become increasingly commercialized. He asserted the following: ‘I predict that the increased fiscal pressure on non-profits will lead them to generate new, more creative forms of commercial activities, and that these new forms will further blur the distinctions between non-profit organizations and private firms’ (p. 403). Tuckman (1998) observed that, under competitive pressure, non-profits might create subsidiaries or joint ventures and make ‘coevolutionary arrangements’, which are profound commercializations for non-profits.

In conclusion, the realization had clearly emerged early on in the study of the in-between space that equating the non-profit sector to the entire civil society could be ‘misleading’ and ‘destructive’ for empirical research (Hall 1987; DiMaggio and Anheier 1990). Alford (1992 p. 40) wrote that ‘the “sector” has no autonomy, and does not mediate between “public” and “private”, … ‘it’ is not homogeneous, but has distinctively different aspects in relationship to the fundamental institutions of capitalism, state and democracy’. Billis (1993, p. 245) asserted that, ‘Yet if by sector (private, public, or voluntary) we intend to describe a homogenous group of organizations that speak and act in unison, then the concept is indeed shaky’. Hall (1992, p. 28) declared, ‘The sector is an artificial construct, not an institutional reality’.

The Lack of an Operationalization Method of the New Paradigm

This section reviews theorizations from the new paradigm spanning from the early era to the most recent. None of these theorizations provides, or is intended to provide, an operationalization method yet, especially for systematic quantitative work.

Early Theorizations

One leading work is Lohmann’s definition of ‘commons’ (1989). Primarily responding to the adoption of the non-distribution constraint as the definitive characteristic of non-profit organizations, commons are positively defined by five characteristics: (1) participation is uncoerced; (2) participation shares a common purpose; (3) participation must have something in common; (4) participation involves mutuality; and (5) social relations must be fair (Lohmann 1989). Commons remind us that there are other organizations in the space that may not be ‘non-profits’: ‘eleemosynary or donative associations, organizations, and groups engaged in unproductive or volunteer labour, whether or not they are incorporated, recognized by the state, tabulated in national data, or made up of paid employees’ (Lohamann 2001, p. 167).

Other early authors also took pains to interpret the interdependence among sectors or to develop typology to understand the process, but they did not go far enough to provide an operationalizable method. For instance, Langton (1987) developed four common anomalies addressing how non-profits internalize characteristics from other sectors: hybridization, functional overlapping, trait absorption, and impingement. Van Til (1988) constructed a ‘tectonic’ map trying to take into account various environmental and underlying factors that influence the voluntary sector. Two of these are ‘climatic factors’—self-absorbed privatism and voluntaristic concern—while the underlying tectonic factors are solidarity of association, bureaucratic control, economic hegemony, and oligarchic control. Alford (1992) defined three categories of non-profits, depending on whether they serve the interests of democracy, capital, or the state. They are popular non-profits (serving democracy), corporate non-profits (associated with capital), and public non-profit organizations (serving the state).

Recent Theorizations

More recently, theoretical frameworks characterized by the blurry sectoral boundary view emerged. These frameworks recognize that the ‘sectors’ can be overlapped or mixed. Some frameworks also provide classifications for hybrid organizations.

The most prominent framework is perhaps the European literature on ‘welfare mix’ and ‘intermediate area’. Although this strand of literature has been in existence since the early 1990s (e.g. Evers 1990, 1995; Evers and Svetlik 1993; Kramer 2000), it did not gain significant visibility in the American-led research field until recently (Evers and Laville 2004). The welfare mix includes the state, market economy, third-sector organizations, and informal communities (private households), which constitutes the sphere that provides welfare services. Evers (1990, 1995) interprets the space (‘the intermediate area’) as a tension field that is influenced by all three sectors. Within the intermediate area, hybrid organizations try to balance values and practices from multiple sectors. A hybrid organization can simultaneously rely on resources from the market, the state, and the community; it can counterbalance both for-profit and non-profit rationales; it can integrate both paid and voluntary work; and it can balance both individual materialistic motivations and collective efforts. This approach disregards the sectoral boundaries and promotes an ‘open, mixed and pluralistic nature of a third sector’ (Evers and Laville 2004, p. 14).

Billis (2010) recognizes the increasing interests in, and lack of a rigorous definition of, hybrid organizations. He hypothesizes that each organization has a primary adherence to the principles of one sector (Billis 1991, 1993, 2010) and that the three sectors overlap. He identifies nine types of organizations, depending on where the organization is situated, in a Venn diagram. The diagram comprises the public sector, the private sector, and the third sector.

Social economy is another promising concept, although the concept has multiple versions in different countries. For example, the social economy concept in Canada also rejects the sectoral concept and the ownership-base:

Social economy is a bridging concept for organizations that have social objectives central to their mission and their practice, and either have explicit economic objectives or generate some economic value through the services they provide and purchases that they undertake’ (Quarter et al. 2009 p. 4).

Quarter et al. (2009) developed four types of social economy organizations. Three types of organizations are situated in the overlapping areas with the public sector and the private sector: community economic development organizations, public sector non-profits, and social economy businesses. Organizations not overlapped with other sectors are dubbed as civil society organizations. Quarter et al.’s typology also employs a Venn diagram that comprised the public, private, and civil society sectors.

Crossan (2007) and Crossan and Van Til (2008) launched a continuum model for organizations in the space. The continuum moves from social activity to economic activity. Four categories of organizations are identified along the continuum: social not-for-profit organizations, social commercial organizations, commercial social organizations, and commercial organizations. The authors claim that the Social Economic Continuum is largely developed to illustrate the location of not-for-profit organizations in relation to commercial enterprises. Working with a four-sector model (the economy, government, civil society, and informal sectors), Van Til (2008) suggests that five other continua can be developed between four sectors in addition to the Social Economic Continuum.

Table 2 provides a developmental map that eventually led to the emergence of the new paradigm. Each framework contributes to the development of theoretical elements of the new paradigm, which is discussed in the following.

Table 2 Systems of classification of the in-between space: the new paradigm
Full size table

The Developmental Map of the New Paradigmatic Work

The emergence of the new paradigm is the result of a gradual development of multiple streams of work. The left column of Table 2 summarizes the theoretical contributions of each work to the new paradigm. In summary, all works, together, contribute to the following characteristics of a new paradigm: (1) there are more than non-profit organizations in the in-between space, and the ownership-based or legal definitions of non-profits are inadequate for sound empirical research of all organizations in the space. (2) In a hybrid organization, multiple dimensions interact and develop different ways of hybridization. (3) Organizations in the in-between space often internalize characteristics from the private and public sectors. (4) The institutional logic lens can assist in developing classifications of hybrid organizations. (5) Provision of social service is not just an effort of the non-profit sector or the government alone, but all four sectors contribute (including households). (6) Different hybrid organizations can have different proportionate mixtures of social and economic activities.

New Paradigm Needs New Operationalization Method

The contribution of this article is to, hopefully, fulfil another urgent task of the new paradigm—developing an operationalization tool for emerging empirical inquiries. This section first presents the challenges associated with developing an operationalization method for the new paradigm. Then, the organizational identity (OI) approach will be introduced to address these challenges.

New Challenges

All the aforementioned recently emerged frameworks are multidimensional, which means they accept that the space embodies characteristics from multiple sectors or of multiple institutional logics (also see Smith 2014). Multidimensionality presents some challenges. First, some frameworks still rely on a three-sector model, but they accept that the sectoral boundaries are more interactive. These frameworks use three ‘dimensions’ to represent three sectors. From an organizational theory perspective, adopting only one dimension (such as the civil society dimension) to represent the in-between space is inappropriate because this space enacts multiple dimensions. Knutsen’s study (2012) reveals that the non-profits in her sample enact at least six kinds of institutional logics: two from the public sector (the logic of the state and democracy), one from the private sector (the logic of capitalism), and three from the in-between space (the logic of religion, profession, and family). Additionally, not all dimension s/institutional logics have to be sector-based. For example, a hospital can have both a caring and a science logic and a university can have both a research and a teaching logic (Dunn and Jones 2010). So the first challenge is to determine what dimensions are definitive for a framework.

The second challenge involves the presence of too many dimensions. If we follow Knutsen’s (2012) findings, there are at least six dimensions in the in-between space. If we consider each logic as a potential dimension for classification, the combinations can give us more than 60 types of organizations. Such classifications are too complicated and defeat the purpose of classification. If so, how can we conceptualize hybrid organizations in the space and then develop tools to measure them?

An Application of Organizational Identity

Organizational ecologists recognize the ambiguity associated with defining organizational forms, particularly for hybrid organizations that embody multiple dimensions. They propose to define organizational form by specifying organizational identity (OI) (Carroll and Swaminathan 2000; Pólos et al. 2002; Ruef 2000; Baron 2004; Hannan et al. 2005; Hsu and Hannan 2005; Hannan et al. 2007). The OI approach circumvents the aforementioned challenges. Instead of conceptualizing hybrid organizations by predetermined classifications (determined by dimensions), the OI approach conceptualizes a hybrid organization according to the perception of selected ‘audiences’. An organization can be conceptualized as different kinds of organizations, depending on what the research question is and who the appropriate audience are. Thus, this approach recognizes that an organization can simultaneously belong to ‘multiple’ categories, but only one matters in a certain research setting and to appropriately selected audiences. The OI approach resolves the challenges by selecting the appropriate audience, and the specific research setting provides idiosyncratic answer to the form of subject organizations.

In this article, the OI approach is introduced primarily to assist with measuring hybrid organizations or other organizations in the in-between space with multiple identities. Our central question is how researchers determine what organizations to be included in the sample when organizations carry multiple identities? For example, if a researcher is studying social enterprises, should he/she include social service non-profits that conduct some business?

The second contribution of the OI approach to the study of the in-between space is its interpretive power. This contribution will be demonstrated throughout the following discussions via examples, such as the definition debates in the field. This contribution is particularly meaningful for organizations in the in-between space, which often have normative and value orientations.

The Logic of the OI Approach

Organizational ecology is dedicated to studying the life histories of all organizations in an organizational field. Therefore, it is critical to define the ‘form’ of organizations in order to draw the boundaries of organizational field and select subject organizations. However, it is acknowledged that ‘form’ has always been an ambiguous concept (Pólos et al. 2002). Conventionally used industrial categories are considered merely as ‘nominal forms’, but not ‘sociologically real forms’ (Hsu and Hannan 2005, p. 478).

The OI approach has been called revolutionary because it leads to a call for building nonmonotonic logic theory, rather than classical first-order logic theory (Hannan et al. 2007). Researchers have conventionally used surface attributes to draw the boundary of a certain organizational form, such as ‘hospitals’ and ‘universities’ (Hsu and Hannan 2005). Logicians identify this convention as relying on so-called ‘first-order logic’, which requires that an organization either belongs to (true) or does not belong (false) to a certain organizational form (Hannan et al. 2007). As such, an organization is either a university or it is not a university, although in reality, some organizations, such as Maharishi Management or National Defense, are only somewhat universities (Hannan et al. 2007). The fact is that an organization may belong to a ‘form’, but at different degrees. Hannan et al. (2007) advocated developing sociological theories that rely on ‘nonmonotonic logic’ to resolve this issue.

Nonmonotonic logic recognizes that a concept is defined by its ‘extension’. For example, all red objects define what ‘red’ is, rather than the reverse: an object is either red or not red. Nonmonotonic logic acknowledges that a social unit is often not clear-cut (‘fuzzy’) and that subjective opinions matter. Applying nonmontonic logic, Maharishi Management and National Defense can have positive but low grades of university membership.

Conceptualizing OI and Organizational Form

The OI approach defines the organizational form by the perception of the agents who have ‘control over material and symbolic resources that affect the success and failure’ of the organization (Hsu and Hannan 2005, p. 476). The collections of these agents are called ‘audiences’. Audiences may give an organization a high or low grade of membership to a certain organizational form. The process whereby audiences determine the membership grade is called ‘valuation’. Audiences uphold ‘default expectations’ about an organization’s properties. These properties are translated into ‘social codes, or sets of rules, specifying the features that an organization is expected to possess’ (p. 475).

There are two steps for audiences to determine the membership of an organization (Phillips and Zuckerman 2001; Hsu and Hannan 2005). First, the audiences screen candidates. Second, the audiences inspect the selected candidates more carefully and develop further refined qualifications of membership, until such qualifications become ‘default’ or ‘taken-for-granted’. By now, the audiences have developed a category or an organizational form with specific qualifications. Then, the audiences keep applying such qualifications to future organizations; these qualifications (‘codes’ and ‘rules’) become the organization’s identity that can be used to determine the membership of an organizational form (Hsu and Hannan 2005, p. 476).

Not everybody who perceives the organization is qualified as an ‘audience’. Audiences contain only the agents who have the power of valuating the organization by rewarding the organization with material and symbolic resources (in cases when the organization’s actions satisfy its social codes upheld by the audience), or devaluating the organization (in cases when the organization fails to satisfy its social codes). For example, donors have it in their mind as to what a charity is, in colloquial terms, regardless of the legal or academic definitions. Donors may sustain the identity code of ‘helping the poor’ to decide which organization is a charity. One day, a charity opens up a franchise restaurant. Donors may not know that the purpose of the restaurant is to subsidize charitable activities, but instead, donors perceive the restaurant as a profit-making enterprise. In this case, donors ‘devalue’ this organization as a charity; consequentially, they stop making donations. However, if the same charity opens up a thrift shop to subsidize charitable activities, donors may consider the thrift shop as conforming to the identity code of ‘helping the poor’ by offering more affordable goods; consequentially, donors maintain their donations. This example illustrates the valuation process and it also demonstrates the interpretive power of the OI approach. It explains why people tend to be more receptive to the idea of non-profits opening thrift stores rather than business ventures. The OI approach may suggest that non-profits need to advertise widely that the purpose of the restaurant is to subsidize charitable activities or should instead avoid the restaurant idea if the major donor’s default identity code of a charity is in conflict with the identity of a restaurant.

Hsu and Hannan (2005) emphasized that forms enforced by audiences in the manner of valuation and devaluation are sociologically real forms. Forms that are defined by surface traits are nominal forms. Canada’s ‘charitable status’ of non-profits is another example of an audience’s valuation process in defining sociologically real form. The Canada Revenue Agency determines whether organizations can earn ‘charitable’ status to become tax-exempt. Charitable non-profits can engage in different charitable activities and they may or may not directly serve the poor. Because of their charitable status, some charitable non-profits tell their clients that the organization is a ‘charity;’ as such, these organizations sometimes experience the loss of clients due to the term ‘charity’. The reason is that these clients often perceive a ‘charity’ as an organization that sustains the code of ‘helping the poor’ by its colloquial meaning, and they do not want to be perceived as poor by receiving services from a ‘charity’. This example illustrates that audiences hold their own default code, regardless the legal definition of a ‘charity’. This example also simultaneously demonstrates the interpretive power of the OI approach, which can explain a phenomenon that is not easily explained by other positivist frameworks. The IO approach may suggest that non-profits should avoid using the term “charity” to avoid potential clients applying the identity code of “helping the poor”, and instead using the term “social service non-profits” to avoid clients’ potential embarrassment. Identity codes can link to normative values and ideals of clients and assist in interpreting norm or value-based behaviour.

Rationale of Applying the OI Approach to In-Between Space Organizations

The OI approach ‘allows for a sounder treatment of ‘hybrid’ organizations, those belonging to multiple classes, categories, or forms’ (Hannan et al. 2007, p. 100). It recognizes that organizations may not squarely fit in a certain form; therefore, form membership is a matter of degree. For instance, a social enterprise can be both a social advocate group and a commercial firm, but it may have a high grade of social group membership and a low grade of commercial firm membership. A hospital may be seen as having a high grade of public organization membership (by the government), a low grade of non-profit membership (by legal professionals), and a low grade of private organization membership (by business donors).

Measuring Identity-Based Organizational Form

How might we apply the OI approach in the study of the in-between space? Researchers can identify subject organizations from the view of selected audiences. This approach requires researchers to select organizations from the point of view of the appropriate audiences. Instead of asking which criteria should be used to define the organizations to be included in the study (such as: what is the definition of a social enterprise?), researchers ask: (1) who should the audiences be and whose valuation matters? (2) How can we reveal the social codes that determine the organization’s identity (Hsu and Hannan 2005)?

Audience Selection

Hsu and Hannan (2005) recognize that organizations can have multiple identities, depending on multiple audiences. Sometimes, these identities can conflict with each other. In their article, they discussed institutional consolidation of multiple identities and their managerial consequences. However, in measuring organizational identity, when researchers need to deal with multiple or conflicting identity, they emphasize that researchers should ‘consider whether the proposed audience has significant social or material control over the relevant outcomes or issues (p. 483)’. The outcomes should be directly related to the subject of the research question. They provided several published examples. A job candidate is an appropriate audience for studying the relationship between organizational identity and its labour pool (Baron 2004). Critics are appropriate audiences for studying how organizational identity might shape market processes (Hsu 2005; Rao et al., 2003). Mohr (1994) selected relief organizations as the audiences to study the ‘moral identities’ of welfare recipients and examine if welfare recipients have similar identities to their relief organizations. Additionally, Knutsen (2010, 2012) trusted two well-experienced practitioners’ views to identify Chinese ethnic non-profits as her sample, when most organizations were reluctant to accept a Chinese identity due to the government’s funding preference of non-ethnic organizations.

It is worth noting the situation when selected individual audiences cannot agree on the identity of an organization. Hsu and Hannan (2005) did not specifically address this issue because the assumption is that different audiences may have different perceptions toward an organization’s form. It is the audience selection to screen the “right” audience’s view. The OI approach seems to assume that there should be no major discrepancy among selected audiences. If there are discrepancies among selected audience, it only shows that the organization has a low membership grade in that particularly organizational category. The reason is that its major stakeholders themselves are confused about their identity. For example, if a researcher decides to ask some selected clients (“the audience”) of a non-profit service organization about its identity, half of the clients say the organization is a government agency, another half says it is a non-profit organization. It means this organization should not be considered as a “sociologically real” non-profit with a high grade of membership, and should not be included in a study about social service non-profit organization, because half of its clients consider it a public agency. Alternatively, the researcher can still include this organization in the study, but noting the organization has a low grade of membership of social service organization (e.g. through a control variable or a special category). One organizational example can be hospitals especially publicly funded hospitals. Although some consider publicly funded hospitals public, others consider them non-profits. Conventionally, hospitals have been frequently treated as in a different category.

The utility of the OI approach underlines the fact that the OI approach helps the researcher to adopt the sociologically real form for sampling rather than using an imprecise nominal form. A nominal form (e.g. legal definition) can conceal the identity of an organization. In Ontario, LHINs (Local Health Integration Network) work with many community non-profits and they govern these organizations on behalf of the government at an arm’s length. LHIN has decision-making authority over policy, contracts, and funding. Community non-profits consistently refer LHIN “the government”, regardless LHIN’s legal status is non-profit. A researcher studying a government funder’s behaviour should include LHINs as one of its ‘public funders’ without being constrained by its legal non-profit status.

Examples of Audience Selection from Our Field

The following examples are about the multiple terminologies of our field. These examples serve two purposes. First, they illustrates that, for scholars of the in-between space, because of their different disciplinary backgrounds, they perceive the identity of the organizations in the space differently. Such difference may have caused confusion in the past; however, this difference should be allowed to exist in the future, because this difference is caused by different perceptions from different audiences. All perceptions are valid to their own audiences. Attempting to unify such a view is inappropriate and fruitless. Second, these examples illustrate the interpretive power of the OI approach and explain that the past confusion is a result of multiple audiences and the solution may not be unifying audience perception but instead capitalizing on such difference for empirical research.

Scott (1999) observed that among different disciplines, different academics perceive non-profits differently:

…‘[n]onprofit’ tends to be the nomenclature of the economists, most of whom have attempted to define this sector and its activity within the theoretical framework of market economics…. ‘[n]ot-for-profit’ is really a further refinement of the economists’ work, it has been employed chiefly by lawyers and accountants seeking to differentiate between mutual benefit associations…. ‘[v]oluntary sector’ is the language of the sociologists, for whom the central defining characteristics of what we are generally calling the non-profit is that the participation in it is not coercive…. ‘[t]hird sector’ or ‘independent sector’ is principally the vocabulary of the political scientists (p. 48).

The reason economists tend to use the term ‘non-profits’ is because the non-distribution constraint matters the most to economic analysis. Similarly, sociologists may prefer to call such organizations voluntary organizations, because sociologists are mostly attracted by the non-coercive participation nature. Therefore, all names are valid but they may refer to somewhat different groups of organizations. Economists should be permitted to study non-profits that are defined by ‘non-distribution constraint’. Sociologists should be permitted to study ‘voluntary organizations’ identified by their own perception. This should not a problem, unless the sociologists are asked to adopt the definition and the group of organizations used by an economist.

The OI approach can interpret the definition debates that infected the field for more than two decades—the controversy surrounding the conventional structural–operational definition of non-profits (Salamon and Anheier 1997). From the OI approach’s perspective, this definition is accurate and applicable if audiences are appropriately selected. However, applying this definition to the wrong audiences would cause a great deal of confusion. In short, the label of “non-profits” coincides with the concept defined by the structural-operational definition if the audiences are primarily US non-profit scholars. For other audiences, non-profits can be viewed in more diverse fashions; therefore, the definition should not apply to them in a unified fashion. This appropriate setting of the conventional definition is described by Smith (2011, p. 30): ‘In the United States, non-profit organizations play a central role in providing key public services, often with government funding’. Non-profit organizations ‘are usually defined by their high level of formality in terms of legal registration, by the preponderance of external funding in their budgets (as opposed to membership support), and by their roles as intermediaries that sit between grassroots constituencies and communities, and government and other agencies’. This description implicitly refers to the idea that audiences of this definition are usually US academics. People who provide definitions are usually academics, especially in the context of a textbook chapter (The Oxford Handbook of Civil Society). Second, Smith’s description articulates the identity codes of such non-profit organizations carry the following codes: (1) ‘the US’, (2) ‘providing public service’, (3) ‘formal legal registration’, (4) ‘large external funding (often from the government)’, and (5) ‘intermediate’.

Hence, great confusion may be generated around the label of ‘non-profits’. For some, such as community grassroots leaders, organizations that do not distribute profit should be considered as ‘non-profits’ (Smith 1997). Typical grassroots organizations are not, in fact, included by the structural-operational non-profit definition, because they are not typically ‘providing social services’, may or may not be ‘legally registered’, do not always have ‘government funding’, and are not necessarily ‘intermediate’. The OI approach can assist to explain such debates in the field and serve to offer a reconciliation between them. Both views are valid for different audiences. The title of this article reflects this view.

Another example is hospitals and universities. Some audiences feel strongly that hospitals and universities are not ‘non-profits’, and should be treated differently (e.g. Hall 2010, p. 33). We can understand this view by adopting the OI approach. The major code distinguishing hospitals and universities from conventional non-profits is that hospitals and universities are not ‘intermediate’ between government and community. The code ‘intermediate’ is not reflected by hospitals and universities; therefore, they should not be included in the conventional structural-operational definition.

The audience selection perspective also helps us to understand that the concept of the in-between space is country specific. The structural-operational definition is mostly applicable to the US setting but it may not apply to other countries because the perception of the in-between space varies among countries. For example, in the UK, recreational organization, private and secondary schools, higher education, trade unions, professional and business associations part of the voluntary sector are “often not thought of as part of the ‘voluntary sector’” (Kendall and Knapp 1996, 24).

Therefore, adopting the OI lens interprets some of the old debates and strengthens our understanding of the classifications in our perception (such as grassroots organizations, universities, and hospitals) but which are neglected in theoretical classification.

Revealing Identity Codes

What is the appropriate identity of an organization? Normally, this is not a difficult task for non-profit research, because the researchers should have already known whether they want to study ‘non-profits’ or ‘social enterprises’. The challenge is more likely to be audience selection rather than a revealing identity code. However, there could be situations that the organizational identity code is not obvious, especially if the study is not about the commonly used category. For example, a researcher may want to reveal subcategories of social enterprises. In this case, revealing various hidden identities can directly lead to the revealing of these subcategories.

Hsu and Hannan (2005) propose two approaches to reveal the social codes of form-specific identity. The first is to utilize the action of external actors who are related to the organization to indirectly reflect an organization’s identity. Ontario’s long-term care for the elderly can serve as an example. One type of organization is referred to as a ‘community care’ organization by the government funders. These organizations provide community-based services to seniors including meal delivery, friendly visiting, transportation, referrals, and telephone reassurance services (Ontario Ministry of Health 1994) in order to help seniors live independently at home. Another type of organization is often referred to as an ‘access centre’. The Community Care Access Centre (CCAC) also helps seniors live in their own homes by providing both in-home care and coordinating community support services. These two types of organizations seem to be similar. In fact, they are quite different for ‘insiders’. Community care organizations were established in the early 1970s and they were funded and administered by the Ministry of Community and Social Services. The Ministry was operating under a ‘wellness’ model that emphasized community-based decision-making, and focused on holistic approaches. Their culture emphasized preventative health care. The CCACs were established in the 1990s and were funded and administered by the Ministry of Health and Long-Term Care. This ministry operates on a ‘medical’ model that emphasizes the power of physicians and episodic care, focusing on individuals rather than the community.Footnote 2 Additionally, CCACs emphasize efficiency, promoting competition for contracting services, and they also emphasize cross-sector service provision under the influence of New Public Management. Conversely, community care organizations tend to exhibit a certain pride in being not-for-profit, and they are often against business-like activities. In this example, ‘community care organization’ and ‘CCAC’ can be used as identity codes to provide classifications for research purposes. The revealing of these identity codes can only be obtained by speaking to or learning from actors who know these organizations.

Alternatively, Hsu and Hannan (2005) suggest examining the perception and beliefs of evaluators (the chosen audiences) directly through semantic analysis of archival documents (e.g. Ruef 2000; Hsu and Podolny 2005). This approach is perhaps too intensive for the purpose of sampling, but it is promising for research where the purpose is to reveal an organizational form. Technically, we can employ semantic analysis to define ambiguous concepts that include non-profit forms of organizations by collecting archival data, such as all academic and professional publications about non-profit organizations within a certain period. Semantic analysis could help us extract the codes or symbols to interpret what academics and practitioners mean by the terms ‘non-profits’ under specified circumstances, including country settings, audience selection, applicable research questions, and so on. These codes could potentially provide a definition of a ‘non-profit organization’ that is a sociologically real form.

Another language-based approach to detect organizations’ identities is through ‘labels’. Hsu and Hannan (2005) assert that labelling is an important step for a certain OI to achieve the taken-for-granted form status. They argue that a label (1) contributes to the legitimacy of a category; (2) demonstrates the homogeneity of a set of organizations; (3) increases the availability of the category to audiences; and (4) facilitates communication about the category. Researchers should pay attention to the emergence of labels. For example, in Canada, people who work with the non-profit sector often refer to a group of organizations as the MASH sector—municipalities (M), academic institutions (A), school boards (S), and health and social service providers (H). Regardless of their legal public or non-profit status, these four types of organizations share certain characteristics (e.g. primarily funded by tax dollars, serve the general public, and are of larger sizes) that potentially attain an organizational form—MASH.

Conclusions and Implications

First, this article recognizes the contribution made by the non-profit paradigm in its effort to fulfil the ‘central task’—delineating the sector, its size, and its composition. The confusion around the traditional structural–operational definition can be clarified if this definition only applies to organizations that fulfil the five identity codes and does not apply to other organizations in the in-between space.,

The second half of the article is dedicated to seeking a potential operationalization tool to identify hybrid organizations in the in-between space, particularly how to identify and sample organizations with multiple and often conflicting identities. The OI approach circumvents the need to identify dimensions of hybrid organizations and develop complicated classification systems. Instead, the IO approaches utilize selected audiences’ views to identity certain forms of organization that are required by the researcher. In addition to its empirical use, the OI approach also demonstrates its interpretive power to phenomena that are not easily explained, such as the definition debate in our field and the examples of donors favouring non-profits opening a thrift shop rather than a businesses.

If we adopt the IO approach in our future empirical studies, researchers will be studying various organizations in the space under the names of such as ‘MASH organizations’, ‘charity’, ‘community care organizations’, in addition to ‘non-profits’ organizations. Names such as community care organizations and MASH organizations can be their own legitimate category for research and they can be selected from the audience’s view. Commonly refereed to categories such as social enterprises can be sampled together from audiences’ view regardless if the organization is incorporated as non-profit or businesses. In the situation of hybridity, if we adopt the OI approach, we would give up the debates such as whether all non-profits are hybrid (Skelcher and Smith 2015), instead, it is up to the selected audience to decide. For example, in order to prevent hybridity legislations, Ontario Non-profit Network officially stated that non-profit organizations are already hybrid organizations because they also earn business income via ‘fee-for-services’; hence, there is no need for legislation. However, the Ministry thinks differently and considers that hybrid organizations are organizations utilizing private business investment (ONN). Theoretical debates on whether all non-profits are hybrid are not theoretically sound and should be avoided.

This article hopes to start a new way of thinking about the definition debate in our field, and hopefully can make contributions to aid future empirical studies. We hope to see more empirical work generated by the new paradigm and link the conceptualization with more empirical studies.